Curve fitting of transcendental equation with lsqcurvefit

I have a transcendental equation of the form

ω is ydata

μ is xdata

x(1) = A1 (unknown to find through curve fit)

x(2)= A5 (unknown to find through curve fit)

Every other parameters are constants.

Stratergy:

1. Assumin both as a xdata (vector).
2. ydata will be zeros of vector.

predictedzeros  = @(x,xdata)-xdata(2)-sqrt(k_val)/sqrt(m_canti_val+x(2)*B_val*Gap_val*L_val*GWM_density(1)+sqrt(2*xdata(1)*GWM_density(1)/xdata(2))*x(1)*B_val*L_val);%x1,x2 are unknown. 
x0 = [1;1.2];  %initial guess
Xdata=GWM_viscosity'; Ydata=GWM_f_out';
zeroes = zeros(1,length(Xdata))';

x = lsqcurvefit(predictedzeros ,x0,[Xdata;Ydata],zeroes)

I am getting the error as given below.

Error using lsqcurvefit
Function value and YDATA sizes are not
equal.

Please let me guide to do the curve fitting of transcendental equation either though this method or any other method.

Thanks.

Answer 1

[Torsten]

Xdata=GWM_viscosity'; 
Ydata=GWM_f_out';
predictedzeros  = @(x) Ydata + sqrt(k_val)./sqrt(m_canti_val+x(2)*B_val*Gap_val*L_val*GWM_density(1)+sqrt(2)*x(1)*B_val*L_val.*sqrt(GWM_density(1)*Xdata./Ydata)) ;
x0 = [1;1.2];  %initial guess
x = lsqnonlin(predictedzeros,x0)